2001:267 - BAYLET, Donegal

This site was discovered by Wm Harte during the 1860s. It lies on the southern shore of Inch Island just to the east of the northern end of the Farland embankment. (It became Harte’s Site 3.) He described it as follows: ‘No. 3 is likewise much destroyed but what is left of it is in better preservation than the others. It is from two to three feet thick and a great deal of what has been removed has been spread over the land immediately joining it.’ He also went on to note that, in the context of Site 3, ‘Stones were found bearing the marks of fire and one lump of cemented oyster shells, too hard to break except with a crowbar which had been, I think made by the action of fire and pressure, for in it were embedded ashes and charcoal.’ Harte also noted the occurrence of a ‘stone axe made of a brownish quartz’. However, it is not entirely certain that the goose bones and ‘fragment of a ruminant, probably Cervus Elaphus’, refer to discoveries from Site 3.

The Baylet site was rediscovered by Mc Naught and Gallagher and in the intervening five years they have noted a serious deterioration in its condition. This would appear to be due to the change in use of the field from sheep to cattle. In this period two stone axes were recovered from the slump along the front of the midden. A significant number of flint tools were recovered from the fields adjacent to the midden; some may have come from it. While these tools were predominantly of later Mesolithic character, some Neolithic and Bronze Age artefacts were also recovered.

At the beginning of 2000, shells were collected from the site for 14C dating: Beta 151926, 6060 ± 80 bp; Beta 151927, 4940 ± 70 bp. Both are corrected from 13C assays for marine effect.

Excavation
The site today appears to be a small portion of a much larger midden and rests on the top of a south-facing slope 2–3m in height with significant quantities of material slumping down from a slight escarpment. This area of slumped material was 12m across from east to west. Below the grassed area behind the escarpment a portion of the midden still appeared to be preserved. Excavation was undertaken to obtain a minimum controlled record of the site in case of further deterioration and to establish how much was still in place. Three areas were chosen.

In the first a vertical cliff had developed. This was up to 0.5m in height and one band of shells was apparent. The area was in an active state of erosion. A small trench, 1.55m by 1.2m, was opened. Under topsoil was a layer of shells, predominantly oyster but including a pocket of periwinkles, and layers of shell in a soil matrix. This layer was up to 0.3m thick and thinned to the south. A portion of what may be a ground point or related implement was associated with this layer. A sample of shells and a hazelnut fragment were removed for 14C dating. Below the shell layer was a layer of orange brown clay containing traces of burning and charcoal. On removal of an area of clay it appeared that a second layer of shells underlay the clay; this lower layer was up to 0.2m thick.

Trench 2 was placed adjacent to a small cliff and was 0.7m by 0.4m. The purpose was to clean the section and establish the depth of the midden. Orange soil with shell, 0.1–0.2m thick, covered orange sandy silt, 0.2m deep. Below this was a thin layer of shells, 0.05–0.15m thick (bird bones were recovered here). This covered a layer of brown orange sand which lay over a layer of shells containing lenses of soil, mostly oysters but with some concentrations of periwinkles and mussels. This was 0.3–0.4m deep. Below this was sterile orange sand.

Trench 3 was placed at the eastern end of the area of slump and initially a 1m-wide trench running downslope was opened. This was 1m by 6.7m but was only explored to any depth in an area 0.5m in width and 5.5m long. This revealed that the midden was perched on the edge of a slope, which even within the trench dropped 1.5m in a length of 4m. In the southern 3m a series of slumped deposits were uncovered. These were excavated to a depth of 0.3m and in places the bottom of the deposits was not reached. One flint blade was found in this area—probably later Mesolithic.

Excavation of the more in situ deposits revealed a layer (1) of soil and shells above (2) a layer of shells with burnt material (in situ?) 50mm deep. Below this was (3) a layer of shells, 50mm deep, which covered (4) a layer of gravel and needle whelks, 0.15–0.2m deep. The next (5) was a layer of shells with concentrations of charcoal and hazelnut shells (one fish bone and one mammal phalanx were recovered and a sample was taken for 14C dating). The final layer (6) was of gravel and oyster shells.

These deposits lay up against a bank with a different sequence: (a) a layer of comminuted shells and soil, (b) a layer of orange brown loam, (c) a layer of oyster shells and (d) orange loam.
Several test-pits to the north of the escarpment revealed what seemed to be traces of the midden some metres to the north, but no traces of the midden were found 10m behind the escarpment. Similarly, 12m to the east of Trench 3 traces of layers similar to Layer 5 were recovered, but further east the nature of the shelly deposits became less clear.

Discussion
One of the most striking features of the excavation was the occurrence of layers of sand, silt or loam within the midden. At this stage the origin of these layers and whether they are a product of one event are unclear. However, certain deposits appeared to be in situ and a worryingly small area may remain intact. Layers 4–6 in Trench 3 may represent the reworking of part of the midden by the sea at a time of higher sea-level, and it is of interest that some of the artefacts found in the vicinity are slightly water-rolled. Preliminary examination of a concretion of the type referred to by Harte, carried out in Newcastle, suggests that they may have been created underwater.

The site appears to represent the remnants of a more substantial midden which accumulated over a period of more than 1000 years, beginning in the Mesolithic and continuing into the Neolithic, although the sandy loam layers could represent a break in occupation. Even from the limited three days of excavation it would appear to be a much richer site than any other known in north-west Ireland.